README

Data for the replication of RADIO PUBLIC SERVICE ANNOUNCEMENTS AND VOTER PARTICIPATION AMONG NATIVE AMERICANS: 
	EVIDENCE FROM TWO FIELD EXPERIMENTS
Eline A. de Rooij and Donald P. Green
Political Behavior (POBE)
Date: July 8, 2016
*****************************************************************************************************************

Files for replication:

1. 	deRooij&Green_POBE2016_replication.do: 					Replication code for Table 1, part of Table 2, Tables A.9-A.12, part of Tables A.13-A.16; Stata format 
2. 	deRooij&Green_POBE2016.dta:  						Replication data; Stata format
3. 	deRooij&Green_POBE2016_replication.R:  					Replication code for part of Table 2, part of Tables A.13-A.16; R format

4. 	deRooij&Green_POBE2016_meta analysis example of bayesian updating.xls: 	Meta analysis example of bayesian updating with Native American radio results (calculations for Figure 1); Excel format
5. 	deRooij&Green_POBE2016_radio_bayes_graph.r:				Replication code for Figure 1; R format
6.	deRooij&Green_POBE2016_TablesA1_A2_A3.do:				Replication code for Tables A.1, A.2 and A.3; Stata format
7. 	deRooij&Green_POBE2016_TableA4.SPS:					Replication code for Table A.4; SPSS format


Software versions:

All data analyses in this article were carried out using Stata/SE 14.1 for Windows (64-bit x86-64)  

R Studio [Version 0.99.879] and R [Version 3.2.3 (2015-12-10) -- "Wooden Christmas-Tree"]

and IBM SPSS Statistics 23 for Windows (64-bit) 


Please execute the code files in the following order:

1.	Execute deRooij&Green_POBE2016_replication.do
2.	Execute deRooij&Green_POBE2016_replication.R
3.	Execute deRooij&Green_POBE2016_radio_bayes_graph.r
4. 	Execute deRooij&Green_POBE2016_TablesA1_A2_A3.do
5. 	Execute deRooij&Green_POBE2016_TableA4.SPS


LIST OF VARIABLES in deRooij&Green_POBE2016.dta:

* Note: variables are numeric and have no missing values, unless otherwise indicated 

NAME			LABEL									VALUES/NOTES
station 		Radio station, string							string variable with 52 unique values
state_id 		State in which radio station mainly broadcasts				1=AK (Alaska); 2=AZ (Arizona); 3=CA (California); 4=CO (Colorado); 5=MN (Minnesota); 6=MT (Montana); 
												7=NM (New Mexico); 8=ND (North Dakota); 9=OK (Oklahoma); 10=OR (Oregon); 11=SD (South Dakota); 
												12=WA (Washington); 13=WI (Wisconsin); 14=WY (Wyoming)
treatment 		Radio station in treatment group 					0=not in treatment group; 1=in treatment group
contact 		Radio station broadcasted registration/GOTV message 			0=did not broadcast message; 1=broadcast message
natam_vf 		Number of Native Americans in propagation zone, voter file		
total_regna 		Number of registered Native Americans in propagation zone, voter file
natam_c2000 		Number of Native Americans in propagation zone, census 2000
pop_c2000 		Total population in propagation zone, census 2000
prop_natam_c2000	Proportion of Native Americans, census 2000 				equals natam_c2000/pop_c2000
v2000_na		Number of Native American voters in 2000 General Presidential election	48/85 missing 
v2004_na 		Number of Native American voters in 2004 General Presidential election	48/85 missing 
v2008_na 		Number of Native American voters in 2008 General Presidential election	 
v2008p_na 		Number of Native American voters in 2008 State Primaries		48/85 missing 
change_0408 		Change in number of Native American voters 2004 to 2008			equals v2008_na-v2004_na; 48/85 missing
change_0008 		Change in number of Native American voters 2000 to 2008			equals v2008_na-v2000_na; 48/85 missing
natam_state 		Sum of Native Americans in propagation zones in same state		equals sum of natam_c2000 within a state in the same experimental round
pop_state 		Sum of total population in propagation zones in same state		equals sum of pop_c2000 within a state in the same experimental round
prnatam_state 		Proportion of Native Americans in propagation zones in same state	equals natam_state/pop_state
exp_round 		Experimental round							2008=experiment conducted during 2008 presidential election; 
												2010=experiment conducted during 2010 midterm election 
treatment_2008		Radio station in treatment group in 2008 experiment			0=not in treatment group in 2008; 1=in treatment group in 2008; 
												52/85 missing (stations in 2008 coded missing, as are new stations in 2010) 
total_reg 		Total number of registered voters in May 2011				37/85 missing
total_reg_dec2011 	Total number of registered voters in Dec 2011				37/85 missing
v2006 			Number of voters in 2006 election					37/85 missing
v2008 			Number of voters in 2008 election					37/85 missing	
v2010 			Number of voters in 2010 election					37/85 missing
v2006_na 		Number of Native American voters in 2006 election			37/85 missing
v2010_na 		Number of Native American voters in 2010 election			37/85 missing
change_0810na 		Change in number of Native American voters 2008 to 2010			37/85 missing
change_0610na 		Change in number of Native American voters 2006 to 2010			37/85 missing
zip 			Number of registration zip codes within propagation zone		37/85 missing
station_id 		Radio station, numeric							numerical version of variable station
natam_station 		Radio station targets mainly NatAm listeners and/or is run by NatAm	0=not a Native American radio station; 1=Native American radio station
outcome 		Change in number of NatAm voters 2008-2004 or 2010-2006			change_0408 and change_0610na combined into one variable
state_yr		State in which radio station mainly broadcasts and experimental round	1=AK08; 2=AZ08; 3=CA08; 4=CO08; 5=MN08; 6=NM08; 7=OK08; 8=OR08; 9=SD08; 
												10=WA08; 11=WI08; 12=WY08; 13=AK10; 14=AZ10; 15=CO10; 16=MN10; 17=MT10; 
												18=NM10; 19=ND10; 20=OK10; 21=OR10; 22=SD10; 23=WA10; 24=WI10; 25=WY10


